This Rust project demonstrates efficient data processing using parallelism. The code leverages Rust's threading capabilities to process data in parallel when the dataset exceeds a specified threshold. It includes both single-threaded and multi-threaded processing based on the input size, making it versatile for various data sizes and system capabilities.
- Dynamic Parallel Processing: Automatically adjusts the number of threads based on the available parallelism on the system.
- Customizable Data Transformation: Supports user-defined data transformation functions.
- Threshold-Based Parallelism: Switches between single-threaded and multi-threaded processing depending on the input size and a user-defined threshold.
| Benchmark | Time per Iteration (ns) |
|---|---|
| bench_heavy_loaded_multi_thread_processing | 254,047.01 ns/iter (+/- 11,405.34) |
| bench_multi_thread_processing | 51,913.97 ns/iter (+/- 1,190.89) |
| bench_single_thread_processing | 75.30 ns/iter (+/- 0.70) |
To run the main program:
cargo run
To run the tests:
cargo test
To run the benchmarks:
cargo bench
This project is licensed under the MIT License.